Investigation of Sludge in Anaerobic Sludge Bed Bioreactor

(1)

University College of Southeast Norway Faculty of Technology Master’s Thesis– Study programme: Energy and Environmental Technology Spring 2016

Hartantyo Seto Guntoro

Investigation of Sludge in Anaerobic Sludge Bed

Bioreactor

(2)

University College of Southeast Norway Faculty of Techology

Department of Process, Energy and Environmental Technology Kjølnes ring 56

3918 Porsgrunn, Norway http://www.usn.no

(3)

MASTER’S THESIS, COURSE CODE FMH606 Student: Hartantyo Seto Guntoro

Thesis title: Investigation of sludge in anaerobic sludge bed bioreactors Signature: . . .

Number of pages: <#>

Keywords: UASB granules Coffee grits Settleability

Supervisor: Rune Bakke Sign.: . . . 2^nd supervisor: Eshetu Janka Wakjera Sign.: . . . Censor: Sign.: . . . External partner: Sign.: . . . Availability: Open

Archive approval (supervisor signature): Sign.: . . . Date : . . . Abstract:

The variation in granular sludge properties from various up-flow anaerobic sludge bed (UASB) reactors and the application of coffee grit as a physical model to characterize granular sludge were investigated in this study. The methods applied for evaluation were density measurements, settling profile and velocity, diameter measurements, total solids and volatile solids, total suspended solids and volatile suspended solids and settleability of granular sludge. Experiments showed that coffee grits can be used as a physical model of granular sludge since it had similar characteristics especially in terms of settling and particle size while less so in terms of density but in a similar range. It was found that a weak correlation (R²= 0.15) between density and settling velocity for coffee grits which might due to the non-uniformity in shape and particle size variation of coffee grits. For the granular sludge, higher density leads to a faster settling velocity and vice versa (R² = 0.73). Moreover, it was found that Saugbrugs (new and old) properties were a slight changed with time, changes that may be good for process performance.

Sample J (E-Convert old) had the highest size of granules (0.44 - 3.91 mm) and sample E (UASB – Econvert) had the lowest organic content which can be seen from its volatile to total solid ratio (0.55).

University College of Southeast Norway accepts no responsibility for results and conclusions presented in this report.

(4)

Preface

This report is performed as a Master Thesis subject FMH606 at University College of Southeast Norway (USN).

I would greatly like to express my gratitude to Rune Bakke and Eshetu Janka, the supervisors of the master thesis. Not only did they cooperated fully by holding a meeting regularly, but they provided useful information regarding this report based on their research experiences and knowledge. I also would like to show much appreciation to my family and friends for support and pray during my study in Norway.

Porsgrunn, June 15 2016 Hartantyo Seto Guntoro

(7)

1 Introduction

1.1 Background

Anaerobic digestion, AD, is a method where micro-organisms mineralize organic matter and generating biogas. Up-flow anaerobic sludge blanket (UASB) reactors are used to obtain high efficiency AD and tested this principle for various applications. Sludge retention time (SRT) is one of determining factors for the process efficiency and this depends on sludge characteristics. The variation in sludge quality become a reason of this topic and learn more about how to maintain adequate sludge quality to avoid problems.

1.2 Problem description

This master thesis project includes theoretical and experimental evaluation of sludge from various up-flow anaerobic sludge bed reactors (UASB). The main goal of this study was to investigate the variations in sludge characteristics. Coffee grits were used as physical model to characterize granular sludge and to test the experimental methods of this particular study was also another aim to conduct this study.

A total of three samples of coffee grits from different brands of coffee and seven samples of granular sludge from various UASB reactors at different periods are used. The list of the samples are in chapter 3 of this thesis. The structure of the report are as follows; chapter 2 describes literature study on UASB and granular sludge, chapter 3 describes material and methods used to investigate the anaerobic sludge characteristics, chapter 4 shows the results from several experiments conducted in the laboratory and chapter 5 describes the discussion of the results by comparing with relevant literatures.

The main granular sludge that were investigated in details were obtained from a full scale UASB wastewater treatment plant saugbrugs reactor at the Saugbrugs factory. The saugbrug UASB reactor is a reactor treating pulp and paper process wastewater at ‘Norske Skog Saugbrugs’ in Halden, Norway [1]. The parameters of saugbrugs reactor has been changed, hence the sludge before and after changing were investigated and the difference of the sludge were examined in this thesis.

(8)

2 Literature study

2.1 Anaerobic treatment

Anaerobic processes have been used for the treatment of domestic and industrial wastewater for over a century. These processes convert organic matter into methane [2]. The decomposition of organic matter occurs in four stages (Figure 2-1): hydrolysis, acidogenesis, acetogenesis and methanogenesis.

Figure 2-1. Degradation steps of anaerobic digestion process [3].

 Hydrolysis

Hydrolysis is the first stage of anaerobic digestion. In this stage, bacteria transform the particulate organic substrate into liquified monomers and polymers i.e. proteins, carbohydrates and fats are transformed to amino acids, monosaccharides and fatty acids respectively [3].

 Acidogenesis

Acidogenesis is the next step of anaerobic digestion after hydrolysis. In this step, acidogenesis microorganism further break down the biomass product after hydrolysis. These fermentative bacteria produce an acidic environment while acetic acids, hydrogen, carbon dioxide, alcohol

(9)

 Acetogenesis

In this third stage, the rest of acidogenesis products i.e. propionic acid, butyric acid and alcohols are transformed by acetogenic bacteria into hydrogen, carbon and acetic acid [3].

 Methanogenesis

During this stage, microorganism converts the acetic acid and hydrogen to methane gas and carbon dioxide. Methanogens are the bactria responsible for this conversion. Waste stabilization is accomplished when carbon dioxide and methane gas are produced [3].

According to Lucas Seghezzo et al. (1998)[4], the advantages of anaerobic sewage treatment are as follow.

1. High efficiency.

Even at high loading rates and low temperature, good removal efficiency can be achieved in the system [4].

2. Low energy consumption.

The energy consumption of the reactor is almost negligible as far as no heating of the influent is needed to reach working temperature and the plant operation can be done by gravity .[4]

3. Low sludge production.

Anaerobic bacteria have slow growth rate. Hence, the sludge production is low compare to aerobic method [4].

4. Low space requirements.

The area needed for the reactor is small when high loading rates are accommodated [4].

5. Low nutrients and chemical requirement.

An adequate and stable pH can be maintained without the addition of chemicals, especially in the case of sewage [4].

6. Simplicity.

The construction and operation of the reactor is relatively simple [4].

7. Flexibility

Can easily be applied on either a very large or a very small scale [4].

Beside the advantages, the anaerobic treatment also has some disadvantages. The disadvantages of anaerobic treatment are as follows [4].

(10)

1. Possible bad odors.

Anaerobic process produces hydrogen sulfide, especially when the influent consist high concentration of sulphate. Hence, proper handling is required to avoid bad odors [4].

2. Long start up.

When no good inoculum is available, the start-up takes longer than aerobic treatment due to methanogenic organisms has low growth rate [4].

3. Low pathogen and nutrient removal.

Pathogen removal is partially only (except helminth eggs, which are effectively captured in the sludge bed) and nutrient removal is not complete and require post treatment [4].

4. Necessity of post treatment.

Post treatment is generally required to reach the standard discharge of organic matters [4].

2.2 UASB reactor

One of the most notable developments in anaerobic treatment process technology was the up- flow anaerobic sludge blanket (UASB) reactor in the late 1970s in the Netherlands by Lettinga and his coworkers [5]. The schematic diagram of a laboratory scale UASB reactor is illustrated on Figure 2-2.

Figure 2-2. Schematic diagram of a laboratory scale UASB reactor [6]

(11)

Influent wastewater is distributed at the bottom of the UASB reactor and travels in an up-flow mode through the sludge blanket. Critical elements of the UASB reactor design are i) the influent distribution system, ii) the gas-solid separator and iii) the effluent withdrawal design [5].

The wastewater comes in contact with the micro-organisms as the wastewater passes through the sludge bed and anaerobic degradation of the wastewater organics occurs. The treated effluent leaves through an outlet at the top of the reactor. Upward hydraulic turbulence caused by produced biogas providing adequate mixing within the system, hence, mechanical mixing is not required. Three phase (i.e. gas-liquid-solid) separator at the top of reactor is to facilitate granule retention [6]. Granules with good settling properties settling back to the granular sludge bed, while flocculated and dispersed bacteria wash out of the reactor with the effluent.

One of the advantage of UASB reactor compared to traditional anaerobic treatment is the ability to retain high biomass concentrations despite the up-flow velocity of the wastewater and the production of biogas. The sludge retention time is almost independent of the hydraulic retention time. Consequently, the reactor can operate at short hydraulic retention times [7].

The important design consideration for UASB reactors are wastewater characteristics in terms of composition and solid contents, volumetric organic load, up-flow velocity, reactor volume, physical features including the influent distribution system and gas collection system [5].

2.3 Granular sludge

There are various types of conglomerates of microbes such as granules, pellets, flocs and flocculent sludge. According to Dolfing (1987) [8], pellets and granules are conglomerates with a dense structure and these conglomerates present a well-defined appearance after settling.

Flocs and flocculent sludge are conglomerates with a loose structure and they form one homogeneous macroscopic layer after settling.

Granular sludge can be described as a spherical biofilm consisting of a densely packed anaerobic microbial consortium [9]. Granules from successful UASB reactor are very compact and have a high settling velocity. As such, they are able to withstand the effect both of the liquid up-flow velocity and hydraulic shear and become concentrated biomass [10]. According to Visser et al. (1991) [11], granule formation is generally thought to be the result of environmental pressures or selection, with any non-granular material being washed out of the reactor.

(12)

The diameter of sludge granules varies from 0.14 to 5 mm depending upon the wastewater used, the operational conditions and the analytical method. The granules shapes are vary depending on the condition of reactor, but they usually have spherical form [7].

Figure 2-3 shows the picture of granular sludge in microscope from one of UASB reactor (A) and the picture of anaerobic granule in scanning electron microscopy (SEM) (B).

Figure 2-3. Sample of granular sludge in microscope in magnification 20x (A); and scanning electron micrograph (SEM) of anaerobic granules in magnification 2900 x (B) [6]

2.3.1 Mechanisms and models for anaerobic granulation

According to Yu Liu et al. (2002)[12], there are mechanisms and models for anaerobic granulation in UASB system to expedite granules development, design and operate granular

A

(13)

local dehydration models) and ii) surface tension model and structural models (i.e. inert matters in inert nuclei model, cation-bridged bacterial aggregates in divalent cation-bridge model, extracellular polymer (ECP) bound bacterial cells in ECP bonding model and Capetown’s model, filamentous bacterial aggregates in spaghetti theory and crystallized nuclei formation model and syntrophic microcolonies in syntrophic microcolonies model) [12].

2.3.2 Anaerobic granulation influencing factors

The long start-up period required for the development of anaerobic granules is one of the major problem encountered with UASB. However, the use of granular sludge from in-operating UASB reactors as the seed material has the advantage of being able to achieve high organics removal within a short start-up period. The information on the major factors influencing anaerobic granules process is essential when researchers are looking for possible strategy for fast production of anaerobic granules. According to Yu Liu et al. (2002)[12], the factors are as follows :

1. Up-flow velocity and hydraulic retention time

It has been observed that the granulation process in an UASB reactor was favored by the combination of high liquid up-flow velocity and short hydraulic retention time (HRT) [13]. A long HRT combined with a low liquid up-flow velocity may allow dispersed bacterial growth and be less favorable for microbe granulation while a short HRT combined with a high liquid up-flow velocity could cause washout of non- competent bacteria in granulation and subsequently promote sludge granulation [12].

2. Organic loading rate

Organic loading rate (OLR) describes the degree of starvation of the microorganisms in a biological system. A low OLR means that the microorganisms are starved in the reactor and a high OLR would ensure a fast microbial growth [12].

3. Characteristics of substrate

The substrate can be roughly classified into high-energy and low-energy feeds based on the free energy of oxidation of organics. High-energy carbohydrate feed during the UASB start-up period can sustain the acidogens and facilitate the formation of extracellular polymer (ECP) [12] .

The complexity of substrate may exert a selection pressure on microbial diversity in anaerobic granules, which consequently influences the formation and microstructure of granules [12].

4. Characteristics of seed sludge

The quality of a particular seed material can be judged in term of ash content, specific

(14)

sewage sludge and surplus sludge from anaerobic treatment plants. Aerobic activated sludge from sewage treatment plant is another type of seed sludge [12].

According to Yu Liu et al. (2002)[12], it might be expected that anaerobic granulation could be expedited simply by manipulating the composition of seed sludge. However, there is still lack of detail guidelines on which species in seed sludge should be a major component and how to manipulate the species in seed sludge.

5. Addition of polymer

One of important factors for the development of granules from non-granular sludge is the presence of nuclei or biocarrier for microbial attachment growth. Synthetic and natural polymers have been used in coagulation/flocculation processes. Chitosan may play a similar role as ECP substances and significantly enhanced the formation of anaerobic granules in UASB-like reactors. Freely moving polymeric chains may build a bridge between cells and can promote the formation of initial microbial nuclei which is the first step of microbial granulation [12].

6. Addition of cations

There is evidence that the presence of cations (positive divalent and trivalent ions) such as Fe²⁺, Fe³⁺, Ca²⁺, and Mg²⁺ could bind to anion (negatively charged cells) to form a microbial nuclei [12].

7. Reactor temperature

An anaerobic system performance is closely related to temperature variation.

Methanogenic bacteria, a core microbial component of UASB granules, grows slowly in wastewater and their generation time range from 3 days at 35 ^oC to as high as 50 days at 10 ^oC [14]. It shows that when the temperature is lower in an anaerobic reactor, the growth of methanogens would be inhibited. Although relatively high temperature encourages the growth of biosolids.

Most full-scale UASB operate in mesophilic range from 22 to 40 ôC, with optimum temperature of 35 ôC and under thermophilic condition, from 50 to 60 ôC or even higher to 70 ôC [12]. However, extreme thermophilic UASB reactors seem not to be beneficial since additional energy is required in order to maintain the reactor temperature [12].

8. Reactor pH

The effect of the reactor pH on anaerobic granulation had been observed by Teo et al.

(2000)[15]. The research showed that the turbidity of granules decreased with the pH increased in a pH range of 8.5 – 11. It indicates that high pH condition would weaken the granular structure. The granular structure was relatively stable in pH range 5.5 to 8.0 and in pH range 5.0 to 3.0 the strength of granule was decreased [12].

(15)

2.3.3 Characterization of granular sludge

Characteristics of anaerobic granules can be determined by observing from its microstructure, methanogenic activity, surface properties, apparent color, density and size and mechanical strength.

1. Microstructure

Scanning electron microscopy (SEM) and transmission electron microscopy (TEM) are often used to examine the surface and internal structure of granules. Fang et al.(1995)[16], MacLeod et al. (1990)[17] and Guiot et al. (1992)[18] had performed intense research on the ultrastructure of UASB. Based on those observation, a multi- layered structural model was proposed with the acidogenic bacteria dominating the outer layer while the inner layer mainly consist of methanotrix-like bacteria.

Uniform structure of granules also found on researched by Grotenhuis et al.

(1991)[19] and Fang et al. (1995)[16] when filamentous microorganisms were predominant on the surface and in the center of the granules and according to Fang et al. (1995)[16], a layered and uniform microstructure would be developed with proteins and carbohydrates as substrates.

2. Methanogenic activity

Methanogenic activity is represented by the activity of methane-producing bacteria and it is defined as the methane production by unit biomass and time or methane production per unit reactor volume and time. This activity more generally determined by using closed bottle test. The methanogenic activity can be used as an indicator of inhibitory effects on anaerobic granules and to evaluate the performance of a system [12].

3. Surface properties

It has been well known that some environmental conditions such as starvation, oxygen level and liquid ionic strength can change hydrophobicity of cell surface [12]. The hydrophobicity of cell surface is an important affinity force in the self-immobilization and attachment of cells.

The strength of granules quantified by turbidity and surface charge of granules has been researched and measured by Quarmby and Forster (1995)[10]. The result suggested that when the surface negative charge increased, the granules tended to become weaker. It can be concluded that the surface characteristics of sludge is closely correlated with the anaerobic granulation.

4. Apparent color

Anaerobic granules have a dark brown or black surface in general. Granules could become lighter with a hollow core and were gray or even white at low organic loading rate (OLR) and liquid up-flow velocity. The gray and white granules were extremely

(16)

The apparent color of granules should be dependent upon chemical composition, microbial and given hydrodynamic conditions. Thus, the changes in granule color may reflect changes in composition and metabolism of granules [12].

5. Density and size

A higher density leads to a faster settling velocity of sludge while geometric size of granules has dual effects on the performance of UASB system. The probability of washout of granules from the system would be increased if the size of granules is too small. On the other hand, an increase in size of granules would reduce the efficiency of mass transfer inside the granule. The resultant size and density of anaerobic granules are dependent on many factors such as OLR, microbial species, hydrodynamic conditions and so on. Medium size of granules with a diameter of 1.0-2.0 mm with narrow size distribution of granules look the most attractive in industrial practice. The relatively high density of individual granules cause them to settle rapidly and good settleability of granules simplify the separation of effluent from the granules and lead to a simple design and operational [12].

6. Mechanical strength

The stability of granules influenced by the strength of granules. Higher strength leads to a more stable and compact structure of granules. Sonication is one of the method to quantify the strength of granules. Quarmby and Forster (1995)[10] reported that turbidity of sonicated granules was linearly related to the applied COD concentration.

A lower COD loading rate would result in higher strength of anaerobic granules and vice versa.

The other method was proposed by Ghangrekar et al.(1996)[21]. The proposed that granular strength could be expressed in term of an integrity coefficient (%). Integrity coefficient is defined as the ratio of solid in the supernatant to the total weight of granular sludge after 5 min of shaking at 200 rev/min on platform shaker. A low integrity coefficient represents granules able to withstand high shear and abrasion.

Thus, the lower integrity coefficient the greater is the strength of granules.

2.4 Particle settling theory

In general, the settling of particles can be analyzed by means of classic laws of sedimentation formed by Newton and Stokes [5]. Newton’s law yields the terminal particle velocity by equating the gravitational force of the particle to the frictional resistance or drag force. The forces are illustrated in Figure 2.4 below.

(17)

Figure 2-4. Illustrating of forces in Newton’s law of settling theory.

The gravitational force is given by

= ( − ) [2-1]

Where = gravitational force (kg m/s²) = density of particle (kg/m³)

= density of water (kg/m³)

= acceleration due to gravity (9.81 m/s²) = volume of particle (m³)

The frictional drag force is given by

= [2-2]

Where = frictional drag force (kg m/s²) = drag coefficient (unitless)

= cross-sectional of projected area of particles in direction of flow (m²)

= density of water (kg/m³)

= particle settling velocity (m/s)

(18)

Equating the gravitational force to the frictional drag force for spherical particles yields Newton’s law :

( ) = ≈ − 1 [2-3]

Where _{( )} = terminal velocity of particle (m/s) = diameter of particle (m)

= specific gravity of the particle

The coefficient of drag takes on different values depending on the flow regime surrounding the particle. The drag coefficient is a function of Reynolds number for particles that are approximately spherical is approximated by equation below [5].

= + + 0.34 [2-4]

While the Reynolds number for settling particles is defined as

= = [2-5]

Where µ = dynamic viscosity (Ns/m²) = kinematic viscosity (m²/s)

For non-spherical particles, equation [2-3] needs to be modified and has been proposed by Gregory et al. (1999)[22]. The equation is as follows.

( ) = _∅ ≈ _∅ − 1 [2-6]

Where ø is a shape factor. The value of the shape factor is 1.0 for spheres, 2.0 for sand grains and up to and greater than 20 for fractal floc.

There are three more or less distinct regions, depending on the Reynold’s number. There are

(19)

2.4.1 Settling in the laminar region

For Reynold’s number less than about 1, the predominant force governing for the settling process is viscosity. Assuming spherical particles, the terminal (settling) velocity of particle equation as below [5].

= ≈ [2-7]

2.4.2 Settling in the transition region

Assuming spherical particles, equation [2-3] can be used to determine settling velocity of particle in the transition region [5].

2.4.3 Settling in the turbulent region

In the turbulent region, the predominant forces are inertial forces. A value of 0.4 is used for the coefficient of drag. If the value of 0.4 is substituted into equation [2-6] for Cd and assume sphere particle, the resulting equation as follow [5].

( )= 3.33 ≈ 3.33 − 1 [2-8]

2.5 Hydrodynamics of UASB reactor

The hydrodynamics characteristics of up-flow anaerobic sludge blanket (UASB) reactors has been investigated by Ren et al. (2009)[23]. The study was set up a number of continuously stirred tank reactors (CSTRs) in series to visualize a UASB reactor. The hydrodynamics of such bioreactor was described with an increasing-sized CSTRs (ISC) model. Another studies on UASB hydrodynamics have shown that they could be well described by multi-CSTR (continuous stirred tank reactor) model with commonly used the equal-sized CSTRs (ESC) model and extended equal-sized CSTRs (EESC) model. The schematic diagrams of those models is shown in Figure 2-5.

(20)

Figure 2-5. A schematic diagram of the model: (A) ESC model; (B) EESC model; (C) ISC model [23].

According to Ren et al. (2009)[23], the dispersion coefficient decreased along the axial of the UASB reactor with gradually increasing tank size in the ISC model. The model validation was using experimental results from both laboratory-scale H²-producing and full-scale CH⁴- producing UASB reactors. The simulation result demonstrated that the ISC was better than the other models. The schematic diagram of the laboratory-scale H²-producing UASB reactor is shown in Figure 2-6.

(A) (B) (C)

(21)

A three-dimensional computational fluid dynamics (CFD) simulation was also performed in the study with an Eulerian-Eulerian three-phase-fluid approach to visualize the phase holdup and to explore the flow patterns in UASB reactors and in terms of the flow pattern and dead zone fractions, the results from CFD simulation and the ISC model were comparable [23]. The solid phase holdup distribution in the UASB reactor is shown in Figure 2-7.

Figure 2-7. Transient model predictions of the laboratory-scale H²-producing UASB reactor at HRT of 4.3 h: (A) sludge volume fraction contours of sludge volume fraction at 0.11, 0.22

(A) (B)

(C)

(22)

3 Materials and methods

There are seven samples of granular sludge which were collected from several different full and laboratory scale wastewater treating reactors to be examined in these experiments. Also three different coffee grit samples were used as a physical model to characterize granular sludge. Sample D (HyVAB – Hyberid Vertical Anaerobic Biofilm Bioreactor) and sample E (UASB E-Convert) were treating wastewater from oil refinery. Sample I and J (E-Convert new and old) were from an IR/IC (internal circulation) reactor treating wastewater from paper mill.

Sample F (EGSB reactor) was a sulphid removal reactor fed with synthetic feed which consist of sodium bicarbonate, sulphide and nitrate solutions. And sample G and H (Saugbrugs new and old) were from reactor treating pulp and paper process wastewater at ‘Norske Skog Saugbrugs’ in Halden, Norway [1]. The sample list and identification are shown in Table 3-1.

Table 3-1 Sample list and identification Sample ID Granular Sludge

A Coffee Grit 1 B Coffee Grit 2 C Coffee Grit 3

D HyVAB

E UASB E-Convert F EGSB Reactor G Saugbrugs (New) H Saugbrugs (Old)

I E-Convert (New) J E-Convert (Old)

3.1 Density measurement

Specific sludge density was measured with the pycnometer method [24]. The weight ( ) and volume of the pycnometer were known and the weight of the pycnometer together with inserted sludge were determined ( + ). The pycnometer was filled with water ( ) and the weight of water ( ′ ) was determined by

= − ( + ) [3-1]

The filled water volume ( ′ ) was obtained as

(23)

′ = [3-2]

The volume of measured sludge ( ) was measured from the difference between the volume of water that fills the empty pycnometer ( ) and the previously determined water volume ( ′ ).

= − = [3-3]

The density of water ( ) is temperature dependent, hence temperature measurement of the water was required.

The granules sludge density ( ) was calculated as

= [3-4]

3.2 Settling profile and settling velocity of granular sludge

A glass cylinder of 0.06 m diameter and 0.425 m depth was used in this experiment to determine the settling profile [25]. A granule was randomly selected and time recorded for each granule and twenty granules were measured for each sample. The time of settling was recorded using stopwatch when the granular sludge was reached at each distance point every 0.1 m.

Figure 3-1 shown the experimental set-up. The fluid used in this experiment was distilled water with temperature measured was 21.5^oC

Figure 3-1. Experimental set-up to determine the settling properties.

The average settling time velocity ( ) for this experiment was calculated through simple

0.0 m 0.1 m 0.2 m

0.3 m 0.4 m

granule

(24)

= _∆ [3-5]

3.3 Granule sludge diameter measurement

The VMS-001 USB Microscope, as shown in Figure 3-2, was used to measure sludge diameter.

Calibration of microscope was done before all the measurements following the calibration and measurement procedure as described in manual and quick start guides provided by manufacturer (appendix 2). Properly stirred samples were conducted to obtain good range of granular size measured. Three measurement for each sample was conducted to obtain measurement statistics.

Figure 3-2. VMS -001 USB Microscope

3.4 Total solids and volatile solids

Total solids (TS) of sample was obtained by separating the solid and liquid phase by evaporation. Percent total solid can be calculated as the ratio between sample weight before the evaporation ( ) and the weight after evaporation ( ) multiply by 100.

% = 100 [3-6]

Evaporation process was done by drying oven with 105 ± 3^oC (Termaks B8133 incubator) for about overnight. The procedure was referred to Laboratory Analytical Procedure by National Renewable Energy Laboratory, U.S Department of Energy [26].

Solid remaining after evaporation were dried, weighed and then ignited at 530 ^oC for 15 minutes. The loss of weight by ignition was a measure of the volatile solids.

(25)

% = 100 [3-7]

% = 100 [3-8]

3.5 Total suspended solid and volatile suspended solid

Total suspended solid (TSS) and volatile suspended solid (VSS) experiments were conducted in accordance with standard methods [27].

3.6 Settleability of sludge

Andras et al. (1989) developed a simple method to evaluate the settling properties of sludge granules by fraction exited under certain up-flow velocities in a fractionating device [28].

Figure 3-3 shows the schematic of the experiment.

Figure 3-3. Up-flow velocity test apparatus.

For the coffee grit experiments, about 1/3 of the glass tube volume filled with the coffee grit.

For the granules, about 5 ml of granules were separated into twelve fractions under up-flow liquid velocities of 1.2, 2.2, 3.2, 4.2, 5.2, 8.2, 10.1, 15.1, 19.7, 44.7, 99.4, 187 mh^-1, respectively

(26)

sludge diameter of each fraction also measured by VMS-001 USB Microscope (Veho, Southampton, UK) and the pictures were taken. TSS and VSS in each fraction were also determined using the method described by Andras et al. (1989) [28].

(27)

4 Results

4.1 Density of samples

Figure 4-1 presents the densities measured experimentally for three different coffee grit (sample A, B and C) and seven different granular sludge (sample D, E, F, G, H, I and J). Coffee grit samples had significantly higher density than all granules investigated. Granule densities varied from 1.01 to 1.09 g/cm³ with sample D (HyVAB) had the lowest and sample J (E- Convert (old) had the highest density.

Figure 4-1. Densities of three coffee grit (A-C) and seven granular sludge (D-J).

The density values were taken three times each samples and average values were calculated.

The density values in the graph were in average. The detail value of density measurement is shown in Appendix 3.

4.2 Settling time profile

Settling time profiles measured for all samples are presented in Figure 4-2. E-Convert (old) granules (sample J) settled fastest (137.8 m/h) and HyVAB (sample D) settled slowest (26.5) while coffee grit samples settled in between granular sludge settling range.

Average settling velocity of each granular sludge was calculated with equation 3-5 and the results are presented in Table 4-1. E-Convert (old) and UASB E-Convert granules had much higher settling velocities (137.8 and 132.6 m/h, respectively). Saugbrugs (old) had the second lowest (26.6 m/h) while the Saugbrugs (new) settled about 10 % faster than the old (29.6 m/h).

(28)

Figure 4-2. Settling time profile of three coffee grits and seven granular sludge.

Table 4-1. The calculated average settling velocity of three coffee grits and seven granular sludge.

Sample ID Granular Sludge Average settling velocity (m/h)

A Coffee Grit 1 31.6

B Coffee Grit 2 34.8

C Coffee Grit 3 55.0

D HyVAB 26.5

E UASB E-Convert 132.6

F EGSB Reactor 57.9

G Saugbrugs (New) 29.6

H Saugbrugs (Old) 26.6

I E-Convert (New) 71.2

J E-Convert (Old) 137.8

4.3 Granular sludge diameter range

The calibration of microscope was done before diameter measuring (Figure 4-3). The calibration method was in accordance with microscope manual procedure (Appendix 2). The result was for one millimeter distance, the microscope measured 0.89 mm. Hence, the correction factor for the microscope diameter measurement was 1.12.

(29)

Figure 4-3. Measurement calibration of VMS – 001 USB Microscope.

Hence, every measured values was multiplied by the correction factor (i.e. 1.12) to obtain the correct value of the measurement.

4.3.1 Sample A – Coffee grit 1

The diameter range for coffee grit 1 was from 0.3 – 1.06 mm (0.27 – 0.95 mm before multiplying with correction factor). Figure 4-4 shows the measuring result.

Figure 4-4. The VMS-001 USB microcope picture and the measured diameter of coffee grit 1 with the range from 0.3 – 1.06 mm.

4.3.2 Sample B – Coffee grit 2

The diameter range for coffee grit 2 was 0.35-1.23 mm (0.32 – 1.1 mm before multiplying with correction factor). Figure 4-5 shows the measuring result.

(30)

4.3.3 Sample C – Coffee grit 3

The diameter range for coffee grit 3 was 0.23-1.23 mm (0.21 – 1.1 mm before multiplying with correction factor). Figure 4-6 shows the measuring result.

4.3.4 Sample D – HyVAB

The diameter range granules sampled from HyVAB reactor was 0.52 – 2.25 mm (0.47 – 2.01 mm before multiplying with correction factor). Figure 4-7 shows the measurement of three samples.

Figure 4-7. The VMS-001 USB microcope picture and the measured diameter of granules sampled from HyVAB reactor with the range from 0.52 – 2.25 mm.

(31)

4.3.5 Sample E – UASB E-Convert

The diameter range for granular sludge sampled from a UASB E-Convert was 0.60 – 2.74 mm (0.54 – 2.44 mm before multiplying with correction factor). Figure 4-8 shows the three-samples measurement result.

Figure 4-8. The VMS-001 USB microscope picture and the measured diameter of granules sampled of UASB – E Convert with the range from 0.6 – 2.74 mm.

4.3.6 Sample F – EGSB reactor

The diameter range for EGSB reactor was 0.41 – 2.32 mm (0.37 – 2.07 mm before multiplying with correction factor). Figure 4-9 shows the measured results.

Figure 4-9. The VMS-001 USB microscope picture and the measured diameter of granules sampled from EGSB reactor with the range from 0.41 – 2.32 mm..

4.3.7 Sample G – Saugbrugs (new)

The diameter range for Saugbrugs (new) was 0.58 – 1.73 mm (0.52 – 1.54 mm before multiplying with correction factor). Figure 4-10 shows the measured results.

(32)

Figure 4-10. The VMS-001 USB microscope picture and the measured diameter of granules sampled from Saugbrugs (new) with the range from 0.58 – 1.73 mm.

4.3.8 Sample H – Saugbrugs (old)

The diameter range for Saugbrugs (old) was 0.22 – 1.35 mm (0.20 – 1.21 mm before multiplying with correction factor). Figure 4-11 shows the measured results.

Figure 4-11. The VMS -001 USB microscope picture and the measured diameter of granules sampled from Saugbrugs (old) with the range from 0.22 – 1.35 mm.

4.3.9 Sample I – E-Convert (new)

The diameter range for E-Convert (new) was 0.20 – 3.75 mm (0.18 – 3.34 mm before multiplying with correction factor). Figure 4-12 shows the measured results.

(33)

Figure 4-12. The VMS -001 USB microscope picture and the measured diameter of granules sampled from E-Convert (new) with the range from 0.2 – 3.75 mm.

4.3.10 Sample J – E-Convert (old)

The diameter range for E-Convert (old) was 0.44 – 3.91 mm (0.4 – 3.48 mm before multiplying with correction factor). Figure 4-13 shows the measured results.

Figure 4-13 The VMS -001 USB microscope picture and the measured diameter of granules sampled from E-Convert (old) with range from 0.44 – 3.91 mm.

4.3.11 Summary of diameter measurements

From diameter measurements, it was found that the coffee grit samples particle diameter (i.e sample A, B and C) were in the range of granular sludge particle diameter. Sample H (Saugbrugs (old)) had the smallest diameter range (0.22 – 1.35 mm) while sample J (E-Convert (old)) had the highest diameter range (0.44 – 3.91 mm).

4.4 Total solids and volatile solids

The summary result of percent total solids, volatile solids and the ratio between volatile solids and total solids (VS/TS) for every samples are presented in Table 4-2. The detail result is presented in Appendix 4.

Coffee grits had the highest organic content (VS/TS) of about 97 % for all three samples. The real granules varied much more, from about 55 % to 86 % organics, with a significant fraction

(34)

Table 4-2 Summary of percent total solids and volatile solids and its ratio of three coffee grits and seven granular sludge.

Sample ID % TS %VS VS/TS

A Coffee Grit 1 40.425 38.863 0.961 B Coffee Grit 2 41.626 40.897 0.983 C Coffee Grit 3 38.638 37.751 0.977

D HyVAB 5.603 4.586 0.819

E UASB E-Convert 13.097 7.188 0.549

F EGSB Reactor 6.943 4.930 0.710

G Saugbrugs (New) 6.355 5.381 0.847 H Saugbrugs (Old) 6.905 5.958 0.863 I E-Convert (New) 10.602 8.186 0.772 J E-Convert (Old) 19.262 12.119 0.629

4.5 Total suspended solid (TSS) and volatile suspended solid (VSS)

These measurements were conducted for Saugbrugs (new) and Saugbrugs (old) granular sludge only (sample G and H) to investigate if VSS/TSS changed with time. Table 4-3 shows the result and the detail is presented in Appendix 5.

The VSS/TSS ratio was lower in the old than in the new sample, so the amount of organic compared to fixed solids increased. The opposite trend was seen when measuring VS/TS (Table 4-2).

Table 4-3 TSS and VSS results of granular sludge samples from Saugbrugs new and old sample.

Sample Sample ID TSS (g/L) VSS (g/L) VSS/TSS

G Saugbrugs (new) 34.5 32.67 0.95

H Saugbrugs (old) 33.67 30.63 0.91

4.6 Settleability

4.6.1 Settleability of coffee grits

The settleability test was started with all coffee grit samples to test and established the method and the appropriate vertical velocities. Afterwards these were used to investigate the real granular sludge (saugbrugs new and saugbrugs old) to know the physical characteristics of the

(35)

From this experiment, settling velocity profile of each particular sample can be depicted by plotting up-flow velocity versus TSS exited from the glass test device. The percentage of total solids (TS) exited versus up-flow velocities were also investigated.

4.6.1.1 Coffee grit 1

Up-flow velocities profile for coffee grit 1 for TSS exited and TS exited in percent (%) is shown in Figure 4-14. The detail of the result is presented in Appendix 6. The highest TSS exited was about 500 mg/L with 8.2 m/h up-flow velocity and the highest TS exited in percent (%) was about 2.6 % with the same up-flow velocity.

Figure 4-14. The TSS exited in mg/L (A) and TS exited in percent (%) (B) for coffee grit 1 as a function of Up-flow velocities.

(36)

4.6.1.2 Coffee grit 2

Up-flow velocities profile for coffee grit 2 for TSS exited and TS exited in percent (%) is shown in Figure 4-15. The detail result is shown in Appendix 6. The highest TSS exited was about 2250 mg/L with 1.2 m/h up-flow velocity and the highest TS exited in percent (%) was about 1.2 % with 2.2 m/h up-flow velocity.

Figure 4-15. The TSS exited in mg/L (A) and TS exited in percent (%) (B) for coffee grit 2 as a function of Up-flow velocities.

(37)

4.6.1.3 Coffee grit 3

Up-flow velocities profile for coffee grit 3 for TSS exited and TS exited in percent (%) is shown in Figure 4-16 below. The detail result is shown in Appendix 6. The highest TSS exited was about 900 mg/L with 2.2 m/h up-flow velocity and the highest TS exited in percent (%) was about 2.6 % with 8.2 m/h up-flow velocity.

Figure 4-16. The TSS exited in mg/L (A) and TS exited in percent (%) (B) for coffee grit 3 as

(38)

4.6.2 Settleability of granular sludge

The settleability experiment for granular sludge was conducted with the same procedure and method used in the experiment for coffee grit. However, the higher up-flow velocities were included to ensure that the sludge was completely exited from the glass test device. The additional higher up-flow velocities which were tested to the granular sludge were 44.8, 99.4 and 187 m h^-1. These flow range were achieved by adjusting and calibrating the water flowing from a tap to the test reactor.

During the test, pictures were also taken for each fraction by microscope (Veho, Southampton, UK) and diameter of exited sludge for each fraction was measured.

4.6.2.1 Saugbrugs (old) – Sample H

Up-flow velocities profile for saugbrugs (old) for TSS exited in percent (%) and cumulative solid loss plot in percent (%) is shown in Figure 4-17. It was shown that about 30 % of TSS exited in up-flow velocities 44.8 and 99.4 m/h (Fig. 4-17A) which made about 66 % and 96 % cumulative TSS exited in those up-flow velocities respectively (Fig. 4-17B).

(39)

4.6.2.2 Saugbrugs (new) – Sample G

Up-flow velocities profile for saugbrugs (new) for TSS exited in percent (%) and cumulative solid loss plot in percent (%) is shown in Figure 4-18. It was shown that about 50 % of TSS exited in up-flow velocity 44.8 m/h (Fig. 4-17A) which made about 82 % cumulative TSS exited in that up-flow velocity (Fig. 4-17B).

Figure 4-18. The TSS exited in percent (%) (A) and cumulative TSS exited in percent (%) (B) for granular sludge from Saugbrugs (new) as a function of Up-flow velocities.

(40)

4.7 Granules size distribution

Diameter measurement of each fraction from settleability experiments was conducted to obtain granules size distribution. The results are shown in the following sections.

4.7.1 Saugbrugs (old) – Sample H

 Up-flow velocity 1.2 m h^-1

The diameter range of granular sludge which were washed out (exited) from the glass test device are 0.06 – 0.19 mm (0.05 – 0.17 mm before multiplying with correction factor). The measured size distribution is shown in Figure 4-19.

Figure 4-19. The granular size distribution of Saugbrugs (old) exited at 1.2 m h^-1 up-flow velocity with range from 0.06 – 0.19 mm.

Figure 4-20. The granular size distribution of Saugbrugs (old) exited at 2.2 mh^-1 up-flow velocity with range from 0.09 – 0.38 mm.

(41)

 Up-flow velocity 3.2 m h-1

(42)

Figure 4-24. The granular size distribution of Saugbrugs (old) exited at 8.2 m h^-1 up-flow

(43)

Figure 4-26. The granular size distribution of Saugbrugs (old) exited at 15.1 m h^-1 up-flow

(44)

(45)

 Up-flow velocity 187 m h^-1

Figure 4-30. The granular size distribution of Saugbrugs (old) exited at 187 m h^-1 up-flow

(46)

4.7.2 Saugbrugs (new) – Sample G

Figure 4-31. The granular size distribution of Saugbrugs (new) exited at 1.2 m h^-1 up-flow velocity with range from 0.01 – 0.21 mm.

(47)

Figure 4-33. The granular size distribution of Saugbrugs (new) exited at 3.2 m h^-1 up-flow velocity with range from 0.21 – 0.38 mm..

Figure 4-34. The granular size distribution of Saugbrugs (new) exited at 4.2 m h^-1 up-flow

(48)

(49)

Figure 4-38. The granular size distribution of Saugbrugs (new) exited at 15.1 m h^-1 up-flow

(50)

(51)

 Up-flow velocity 187 m h^-1

Figure 4-42. The granular size distribution of Saugbrugs (new) exited at 187 m h^-1 up-flow

(52)

4.7.3 Summary of Saugbrugs (old and new) granules size distribution

The summary of saugbrugs (old and new) granules size distribution is shown in Table 4-4. The trend shows that higher up-flow velocity leads to bigger diameter of granules which were exited from glass test device for both granules sludge samples (Saugbrugs old and new).

Table 4-4. The summary of Saugbrugs (old and new) granules size distribution according to up-flow velocity.

Up-flow Velocity (m/h)

Size distribution (mm) Saugbrugs

(old) Saugbrugs (new) 1.2 0.06 - 0.19 0.01 - 0.21 2.2 0.09 - 0.38 0.13 - 0.35 3.2 0.06 - 0.20 0.21 - 0.38

4.2 0.1 -0.25 0.26 - 0.6

5.2 0.14 - 0.32 0.19 - 0.38 8.2 0.22 - 0.46 0.22 - 0.49 10.1 0.34 - 0.74 0.16 - 0.44 15.1 0.31 - 0.64 0.26 - 0.66 19.7 0.38 - 0.77 0.55 - 0.79 44.7 0.46 - 1.2 0.79 - 1.48 99.4 0.74 - 2.12 0.85 - 1.34 187 0.97 - 3.00 1.17 - 1.71

(53)

5 Discussion

5.1 Relation between density and settling velocity

According to Figure 4-1 in chapter 4, the different coffee grit had densities higher than but rather close to the density of granular sludge. In fact, this was one of the main reason that this study has used coffee grits as a physical model to characterize the granular sludge as the availability of adequate and different granular sludge samples were limited.

The relation between density and settling velocity for every samples is shown in Figure 5-1. It can be seen that even though the coffee grit samples had higher densities, but they did not have faster settling velocities compare to real granular sludge (sample D, G and H) which had lower densities than coffee grits.

In order to observe and discuss the relation between density of the different samples and the setting velocity, the coffee grit samples (sample A, B and C) and granular sludge samples (sample D, E, F, G, H, I and J) were separated in this discussion. Provided that coffee grit and granular sludge have different characteristics.

Figure 5-1. Measured density of three coffee grits (sample A-C) and seven granular sludge (sample D-J) along with the settling velocity for the same samples.

(54)

Figure 5-2 The density range of three coffee grits along with the settling velocity.

The study has shown that there was weak correlation between density and settling velocity for all coffee grits tested (Figure 5-3). The observed coefficient of determination (R²) between density and settling velocity was 0.15.

Figure 5-3. Correlation of density with average settling velocity for three coffee grits. The error bar represents the standard error and n=3.

It is known that the shape of coffee grit particles were not sphere and the shape were un-uniform as seen in Figure 4-4 until 4-6. According to Gregory et al.(1999)[22], the shape factor (ø) and

(55)

Thus, the un-uniformity of shape and the particle size variation of coffee grits made a weak correlation between density and settling velocity.

5.1.2 Granular sludge

For the granular sludge, the graph of density and its settling velocity is shown in Figure 5-4.

The samples are ordered from higher to lower density. Hence, the graph shows that the higher density of granules, the higher average settling velocity and vice versa. Sample J (E-Convert old) had the highest density and settling velocity (1.09 g/cm³ and 137.8 m/h) while sample D (HyVAB) had the lowest density and settling velocity (1.01 g/cm³ and 26.6 m/h) among other granular sludge samples.

Figure 5-4. The density for seven granular sludge along with the settling velocity.

A strong correlation (R² = 0.73) between the granules density ( ) and the average settling velocity ( ̅) have been found in the study. The equation 5-1 depicts the correlation and figure 5-5 shows the correlation graph.

= 0.0005 ̅ + 0.9989 [5-1]

This correlation is in accordance with Yu Liu et al. (2002)[12] that a higher density leads to a faster settling velocity of sludge.

(56)

Figure 5-5 Correlation of density with average settling velocity for seven granular sludge. The error bar represents the standard error and n=3.

5.2 Relation between density and VS/TS

The ratio between volatile to total solid (VS/TS) is an indicator of the organic fraction in the sludge solids and it is also a good indicator of its level of digestion. For undigested sludge, VS/TS ratio ranges from 0.75 to 0.80, whereas for digested sludge the range is from 0.60 to 0.65 [29].

According to Table 4-2 on chapter 4 above, coffee grit samples (sample A, B and C) have higher VS/TS value than granular sludge samples (Sample D, E, F, G, H, I and J). Sample A, B and C have volatile to total solid ratio 0.96, 0.98 and 0.98 respectively. It means that over 95% of solids are organic solids. In fact, it obvious that the coffee grits are organic matters and undigested.

For the granular sludge, the range of volatile to total solids ratio is from 0.55 to 0.86. The lower of the ratio means the lower of volatile solid content compared to its total solids. On the other words, the sludge has lower organic solid compare to its total solids. Extracellular polymeric (ECP) substances are known to make up to 70% of the organic matter in sludge. The ECP is considered to be responsible for the sludge’s poor dewaterability due to its high water retention capacity [30]. Thus, according to H Saveyn et al. (2009) [30], the lower organic content in the sludge leads to good dewaterability of sludge.

(57)

The statistical relation between coffee grit density and its VS/TS according to the experiments result has shown a week correlation (R² = 0.49).

Figure 5-6. The correlation of density with the ratio of VS/TS for coffee grit. The error bar represents the standard error and n = 3.

The same weak correlation (R² = 0.47) also occurred for density and VS/TS of the granular sludge (Figure 5-7). However, the correlation between the density of coffee grit and granular sludge with its VS/TS were different. Coffee grit had a positive correlation while granular sludge had negative correlation.

Investigation of Sludge in Anaerobic Sludge Bed Bioreactor

Hartantyo Seto Guntoro